High albumin sensitization drives uninterpretable basophil activation tests in cow's milk allergy

The basophil activation test (BAT) is increasingly used for diagnosing IgE-mediated food allergies, yet its performance in cow's milk allergy (CMA) remain insufficiently characterized in real-world pediatric practice. This study evaluates the diagnostic accuracy of BAT for CMA and identifies factors associated with non-interpretable results.

We carried out a single-center retrospective observational study in 59 children with a documented history of cow's milk reactions. Clinical status at the time of BAT was determined using oral food challenges (OFCs), skin prick tests, and specific IgE (sIgE) levels. ROC analyses were conducted to assess the diagnostic performance of BAT parameters and sIgE levels.

We analyzed 65 BAT evaluations performed in 59 children. Only 41 of 65 BATs were interpretable. Among these, BAT showed excellent diagnostic accuracy, with AUROCs of 0.89 for CDmax and 0.89 for BAT-AUC, outperforming sIgE for predicting milk-tolerant or milk-reactive patient. However, 37 % of BATs were non-contributive due to elevated spontaneous activation, affecting 73 % of clinically reactive children. Non-contributive BATs were associated with elevated bovine serum albumin (BSA)-sIgE levels and β-lactoglobulin-sIgE. BSA-sIgE demonstrated the highest ability to predict BAT interpretability (AUROC 0.91). A threshold <0.25 kU_A/L identified patients likely to have an interpretable BAT (NPV 95 %), whereas levels >10.8 kU_A/L predicted BAT failure with 100 % PPV.

BAT has high diagnostic accuracy for CMA when interpretable but is frequently limited by spontaneous activation in highly sensitized children. sIgE thresholds - particularly BSA-sIgE - may help identify patients in whom BAT can be reliably used, supporting more targeted integration of BAT into clinical practice.